The objective of this study was to determine possible differences in the outcome of patients with rheumatoid arthritis (RA) with disease onset early and late in life. As part of a broader outcome study of RA which included patients seen in the division of Rheumatology of Hospital Universitario Marqués de Valdecilla of Santander, Cantabria (Northern Spain) with disease duration between 2 and 7 years, we selected patients with an age at disease onset of <or=45 or >or=65 years. The medical records of all eligible patients were reviewed for relevant clinical and laboratory variables; the patients were then further evaluated for disease activity using biological tests and joint indices such as joint counts and Thompson's Index, functional capacity using the American College of Rheumatology (ACR) functional classification (ACR-FC) and the modified Health Assessment Questionnaire (M-HAQ), and anatomical damage using the number of joint damage (NJD) and radiographs read by the Sharp's scoring method for joint erosion (JE), joint narrowing (JN), and overall. Patients in both subsets were then compared. For the multivariable analyses all patients in the original larger cohort were included, so that age could be used as a continuous variable and the power of the analyses could increase; 31 younger (mean age+/-SD: 36+/-7 years) and 35 older (73+/-6 years) patients were available for assessment. No differences in disease duration and gender distribution were observed. Likewise, both subsets had similar levels of disease activity, both articular indices, and biological markers. In contrast, elderly patients showed more functional limitations as per the M-HAQ [median (interquartile range): 0.4 (0.13-1.2) vs 0.13 (0-0.6), p=0.007] and greater anatomical damage as per the NJD [median (interquartile range): 2 (0-4) and 0 (0-2), respectively, p=0.04] and the JE, JN, and total Sharp Index score (p=0.001, 0.02, and 0.001, respectively). Although older patients took fewer disease-modifying antirheumatic drugs (DMARD) and combined DMARD treatments (2.5+/-1.4 vs 1.9+/-1.3, p=0.05 and 0.8+/-1.1 vs 0.3+/-0.6, p=0.01, respectively), multivariable analysis demonstrated an independent association between age at disease onset and the number of DMARD and functional and anatomical decline. Late-onset RA does not present a better prognosis than the early-onset form of the disease. At the very least the disease is comparable between both patient groups. However, disease compounded by age-associated factors may in fact have a worse prognosis late than early in life.